Chiseling Out Efficiency: Structured Skeleton Supervision for Efficient Code Generation

2026-06-04

AI Summary

This work addresses the performance inefficiency of code generated by large language models, which often stems from the absence of high-efficiency structural patterns. To tackle this issue, the authors propose EffiSkel, a novel framework that explicitly models and learns efficiency-oriented, reusable abstract structural skeletons. By integrating skeleton extraction, structured supervision, and fine-tuning of large language models, EffiSkel jointly optimizes code generation and skeleton prediction through multi-task learning, thereby enhancing execution efficiency while preserving semantic correctness. Evaluated on the Mercury benchmark, EffiSkel significantly outperforms prior methods EffiCoder and CodeDPO, achieving an 11.11% and 3.71% improvement in Efficiency Ratio, and a 0.36 and 0.22 increase in Average Speedup, respectively.

Abstract

Large Language Models (LLMs) are capable of generating syntactically correct and functionally complete programs, greatly streamlining software development. However, recent studies reveal that these programs typically execute substantially slower than human-optimized counterparts. Existing approaches to bridging this efficiency gap typically involve either iteratively optimizing code after generation or fine-tuning models on corpora of efficient code. Yet, these methods expose the model to efficiency signals only by mimicking complete, optimized solutions, without explicitly encoding the structural code patterns essential for achieving high runtime performance. Addressing this gap presents two core challenges: (1) extracting and representing latent, efficiency-oriented structural patterns embedded within complex syntax and control flows, and (2) effectively learning these patterns without destabilizing the semantic training of LLMs. To tackle these challenges, we propose EffiSkel, an efficiency skeleton-guided framework that explicitly extracts and learns efficiency skeletons-abstract, reusable structural patterns underpinning efficient code-by leveraging three complementary strategies. These skeletons are integrated into a multi-task learning regime that jointly optimizes code generation and skeleton prediction. Experiments across multiple programming languages and benchmarks demonstrate that EffiSkel significantly enhances both functional correctness and efficiency, resulting on Mercury with DeepSeek-Coder (7B) a +11.11% (vs. EffiCoder) and +3.71% (vs. CodeDPO) higher Efficiency Ratio (ER), and a +0.36 (vs. EffiCoder) and +0.22 (vs. CodeDPO) increase in Average Speedup (AS). These results highlight the effectiveness of explicitly modeling efficiency skeletons in improving the runtime performance of code generated by LLMs.

Problem

• code efficiency
• large language models
• structural patterns
• runtime performance
• code generation

Innovation

• efficiency skeleton
• structured supervision
• multi-task learning
• code generation
• runtime performance